See below for some new work out of the lab over the last several months.
Michael Ripperger, application developer in the lab, led External Validation and Updating of a Statistical Civilian-Based Suicide Risk Model in US Naval Primary Care. In this cohort study of active-duty Navy service members, we retrained and updated a civilian suicide attempt risk model with DoD-specific variables with improved performance. Domain and temporal validation results were similar to external validation, suggesting that implementing an external model in US Navy primary care clinics may bypass the need for costly internal development and expedite the automation of suicide prevention in these clinics.
Barrett Jones, a recent PhD graduate from the lab, published Sequential autoencoders for feature engineering and pretraining in major depressive disorder risk prediction.
Drew Wilimitis, former statistical analyst in the lab and now a VUMC DBMI PhD student, and Dr. Walsh published a tutorial in JMIR, Practical Considerations and Applied Examples of Cross-Validation for Model Development and Evaluation in Health Care.
Work from a collaboration with Veterans Affairs Rocky Mountain Mental Illness Research Education and Clinical Center, Boston University School of Public Health, Brandeis University, and University of Colorado which aims to improve upon suicide risk prevention strategies for combat Veterans:
- Time-dependent suicide rates among Army soldiers returning from an Afghanistan/Iraq deployment, by military rank and component
- Associations of Military-Related Traumatic Brain Injury With New-Onset Mental Health Conditions and Suicide Risk
- Trends in suicide rates by race and ethnicity among members of the United States Army
- Divergent trends in accidental deaths since return from an Afghanistan/Iraq deployment among army soldiers
Work with the PsycheMERGE Consortium, a partnership of researchers across the world who are working to better understand and treat neuropsychiatric illnesses: